INTRODUCTION
The invasion of Red Sea species into the eastern basin of the Mediterranean through the Suez Canal is known as ‘lessepsian migration’ (Por, Reference Por1978). According to Zenetos et al. (Reference Zenetos, Gofas, Verlaque, Cinar, Garcia Raso, Bianchi, Morri, Azzurro, Bilecenoglu, Froglia, Siokou, Violanti, Sfriso, San Martin, Giangrande, Katagan, Ballesteros, Ramos Espla, Mastrototaro, Ocana, Zingone, Gambi and Streftaris2010), who give an updated list of alien marine species recorded to date from the Mediterranean Sea, a total of 955 alien marine species have been recorded from the Mediterranean, the vast majority of which (718) having being introduced in the Eastern Mediterranean and are thus, presumably, of lessepsian origin.
The taxonomy of the species discussed in this paper is in a state of flux, as until more genetic studies become available it is difficult to group or split the several Brachidontes morphological variants into one or more valid taxa.
On one hand, several geographical forms including Brachidontes pharaonis (Fischer, 1870) are considered by some authors (Lamy, Reference Lamy1936; Huber, Reference Huber2010) to be variants of Brachidontes ustulatus (Lamarck, 1819). This is an Indo-Pacific mytilid with a highly variable shell first described as Mytilus ustulatus, with Lamarck's locus typicus for the type series being Brazil. Lamarck's description was based on specimens collected by the French explorer Nicolas Baudin, presumably during his voyage to Australia spanning from 1800 to 1803. For this reason, Lamy (Reference Lamy1936) argues that the locus typicus is incorrect and the type series is actually from the Indo-Pacific area. According to Huber (Reference Huber2010), the variability of this species has led to the propagation of several synonyms such as Brachidontes semistriatus Krauss, Reference Krauss1848, Mytilus variabilis Krauss, Reference Krauss1848, (and hence Brachidontes variabilis), Mytilus pharaonis Fischer, 1870 and Mytilus arabicus Jousseaume in Lamy (Reference Lamy1919). Lamy reassessed his 1919 opinion in 1936, synonymizing M. arabicus, M. pharaonis, and M. variabilis with their senior synonym B. ustulatus. Huber (Reference Huber2010) reiterates Lamy's view and groups several forms into one vast ‘ustulatus-complex' spreading from the Red Sea (and the westward invasion, cf. Sarà et al., Reference Sarà, Romano, Caruso and Mazzola2000) to Japan, individuals of which complex show a high degree of morphological plasticity.
On the other hand, molecular studies like that of Sirna Terranova et al. (Reference Sirna Terranova, Lo Brutto, Arculeo and Mitton2007) discover genetic divergence between separate geographical populations that are ‘substantially higher' than those between species of other congeners of other mytilid genera such as Mytilus Linné, 1758. According to Sirna Terranova et al. (Reference Sirna Terranova, Lo Brutto, Arculeo and Mitton2007), B. variabilis actually consists of three species—B. pharaonis ‘sensu lato’ (to which the Red Sea and Mediterranean populations belong), an Indian Ocean B. variabilis and a Pacific Ocean B. ‘variabilis'.
In this paper, the authors adhere to the name Brachidontes pharaonis (Fischer, 1870) for the sake of geographical specificity and clarity.
One of the earliest recorded and most successful lessepsian migrants in the Mediterranean Sea is indeed B. pharaonis (cf. Zenetos et al., Reference Zenetos, Gofas, Verlaque, Cinar, Garcia Raso, Bianchi, Morri, Azzurro, Bilecenoglu, Froglia, Siokou, Violanti, Sfriso, San Martin, Giangrande, Katagan, Ballesteros, Ramos Espla, Mastrototaro, Ocana, Zingone, Gambi and Streftaris2010), having established large and stable populations within the basin. The species hails from the Indo-Pacific, and by 1976 it spread as far west as the Red Sea forming mytilid clusters (Gilboa, Reference Gilboa1976; Safriel et al., Reference Safriel, Gilboa and Felsenburg1980). The bivalve was first recorded in the Mediterranean Sea at Port Said just seven years after the opening of the Suez Canal in 1869, and has to date spread from the Israeli coastline, where a study covering the extent of colonization was carried out by Rilov et al. (Reference Rilov, Benayahu and Gasith2004), to the western coast of Sicily (Sarà et al., Reference Sarà, Romano and Mazzola2008), establishing stable populations even in the north Adriatic (De Min & Vio, Reference De Min and Vio1997). Records from the Aegean Sea are sporadic (e.g. Tenekides, Reference Tenekides1989; Dogan et al., Reference Dogan, Onen and Ozturk2007). Sarà et al. (Reference Sarà, Romano, Caruso and Mazzola2000) have predicted that the species will continue its westward spread from Sicily to North Africa and Gibraltar.
The first publications mentioning Brachidontes pharaonis for the Maltese Islands (as Brachidontes variabilis (Krauss)) were those by Cachia (Reference Cachia1975) and Lanfranco (Reference Lanfranco1975); however, records of the species from October 1970, in Qawra and Baħar iċ-Ċagħaq, exist, together with 1973 records from St Julian's, Kalkara, St Thomas' Bay and Birżebbuġa Bay (C. Cachia, personal communication). Most of these rare finds consisted of disarticulated valves, others of which occurred in 1985 in Marsaskala Bay (C. Mifsud, personal communication) (see Table 1 and Figure 1 for historical and unpublished records). Subsequent checklists (Cachia et al., Reference Cachia, Mifsud and Sammut1993; Cachia, Reference Cachia1999; Cachia et al., Reference Cachia, Mifsud and Sammut2004; Sciberras & Schembri, Reference Sciberras and Schembri2007) mention the species with the taxon now amended to Brachidontes pharaonis (P. Fischer, 1870). In Cachia et al. (Reference Cachia, Mifsud and Sammut2004) it is noted that the species occurs as ‘few individuals living at the lower mediolittoral, often embedded in moss-like algae'. Mifsud & Cilia (Reference Mifsud and Cilia2009) reported the first significant colonies with a mytilid cluster habitus from Birżebbuġa.
Previous and current records, with the latter including population densities, for Brachidontes pharaonis (Fischer, 1870) around the Maltese Islands. Named coastal zones indicate embayments where the highest densities of B. pharaonis were recorded or where interesting observations about the species were made. Refer to Table 1 for the numbered records.
Historical and unpublished records of Brachidontes pharaonis (Fischer, 1870) in the Maltese Islands.
Brachidontes pharaonis displays several features that are typical of invading species (Sarà et al., Reference Sarà, Vizzini and Mazzola2003), such as high fecundity and tolerance to variable environmental conditions, including high levels of pollution, according to Morton (Reference Morton1988) and Shefer (Reference Shefer2003). The proliferation of B. pharaonis within the Mediterranean Sea, the ease with which it can be sampled and the evident displacement of indigenous species of Mytilaster Monterosato, 1883 it causes (Felsenburg & Safriel, Reference Felsenburg and Safriel1974; Safriel & Sasson-Frostig, Reference Safriel and Sasson-Frostig1988; Rilov et al., Reference Rilov, Gasith and Benayahu2002) have fuelled physiological (e. g. Sarà, Reference Sarà2006) and even genetic (Shefer et al., Reference Shefer, Abelson, Mokady and Geffen2004; Sirna Terranova et al., Reference Sirna Terranova, Lo Brutto, Arculeo and Mitton2006, Reference Sirna Terranova, Lo Brutto, Arculeo and Mitton2007) studies on the species. The latter investigations have disclosed a lack of genetic structure within the Mediterranean populations for the species, suggesting a relative lack of genetic differentiation between such populations.
The mapping of a species' distribution and the documentation of shifts in such an occurrence have fascinated biologists for decades, ever more so with the ascent of invasion biology. This study represents one of the first comprehensive attempts at mapping the distribution of a non-indigenous species (B. pharaonis) conducted over the entire extent of a national territory—that of the Maltese Islands—with one other similar study being that of Rilov et al. (Reference Rilov, Benayahu and Gasith2004), conducted in Israel. Supplementary information on the density of the mytilid and on the ecological and physical characteristics of each of the sampling sites is also compiled.
MATERIALS AND METHODS
A series of observations on local shores was carried out between 2009 and 2011 by the authors. A stretch of coastline at a time was explored and the presence or absence of B. pharaonis was noted. The entire extent of the low-lying coastline of the Maltese archipelago was investigated in this way, with only coastal stretches having a significant slope value (>10 degrees; e.g. the cliffs along the western coastline of Malta and Gozo) not being sampled.
In the case of positive results, a measurement of the population density was carried out using three randomly-positioned 0.01 m2 replicates, about 1 m above mean sea level (as demarcated by Cystoseira spp. populations), with subsequent extrapolation to a value for one square metre. Locations with very small, scattered numbers of specimens were listed as containing ‘isolated individuals’. The mytilids were not scraped off the rocks prior to counting, therefore the densities observed and reported in this study are an under-estimate, consisting exclusively of superficially visible adults. Using on-site observations and maps (Government of Malta, 1993; Google Maps, Reference Google2011) the type of substratum colonized by B. pharaonis was recorded and coordinates of each sampling location were extracted. Predominant supralittoral to sublittoral macrofloral and macrofaunal species, in close proximity to the B. pharaonis populations or specimens, were identified, at least to generic level.
RESULTS
Figure 1 gives the current distribution of B. pharaonis in the Maltese Islands as reported by the present study. Table 1 lists all the previous records of B. pharaonis from the Maltese Islands, whilst Table 2 lists the findings at each individual sampling site adopted in the present study.
Populations of Brachidontes pharaonis (Fischer, 1870) in the Maltese Islands as discovered during the present research. Records are sorted clockwise beginning from Dwejra, Gozo.
DISCUSSION
Findings from the present study seem to suggest that B. pharaonis has a predilection for sedimentary substrates (limestone), in particular Globigerina Limestone, on which it reaches the highest densities. In fact, the species occurs only as small clusters or as isolated individuals on other types of limestone, such as Upper Coralline or Lower Coralline Limestone. Besides limestone substrates, B. pharaonis exhibited a high degree of environmental plasticity in that it was also recorded in the present survey from concrete, wood, a discarded seaborne tyre and even from within vermetid (Dendropoma petraeum (Monterosato, 1884) and Vermetus sp.) reefs. In the westernmost record for the island of Malta, within the Qarraba area (Figure 1), the species was even observed inhabiting empty boreholes of Lithophaga lithophaga (Linné, 1758).
Although surveys conducted by malacologists since 1970 in the Maltese Islands were not systematic ones but just snapshot studies restricted to point locations only, such that B. pharaonis might have been under-reported in the past, findings of the present study suggest that this species has greatly expanded its range in the archipelago in recent years, reaching, in some areas, the same densities reported, for example, from a hyperhaline saltpan in western Sicily (Sarà, Reference Sarà2006; Sarà et al., Reference Sarà, Romano and Mazzola2008).
The first record of B. pharaonis in the Maltese Islands was at Qawra (point 1 in Figure 1), made in 1970—since then, the species has expanded its range within the archipelago, being recorded also in small densities from the islands of Comino and Gozo (Figure 1). On these two islands, B. pharaonis is restricted mainly to concrete jetties, suggesting that vessels are acting as vectors, at least in initial stages, in proliferating the species.
Displacement of the native Mediterranean mytilid Mytilaster minimus (Poli, 1795) is an often-cited effect of invasion of B. pharaonis (e.g. Safriel & Sasson-Frostig, Reference Safriel and Sasson-Frostig1988; Rilov et al., Reference Rilov, Benayahu and Gasith2004). In many of the sampled localities, an absence of M. minimus was noted; in a small number of localities, like Baħar iċ-Ċagħaq, Balluta and some areas in Valletta (Table 2), a degree of sympatry betwen the two species was observed; however, the density of M. minimus was always much lower than that of the invasive species and on no occasion was a dense cluster of M. minimus observed. It is impossible to accurately define the extent of displacement of B. pharaonis over M. minimus simply because there are no previous studies on the latter's population densities on the Maltese coastline. However, shell sizes exceeding the 10 mm normally attained by M. minimus were very rarely observed in areas where both species were present. Safriel & Sasson-Frostig (Reference Safriel and Sasson-Frostig1988) hypothesize that the two species are able to coexist due to different environmental preferences and an extinction–immigration continuum. Juvenile specimens of another mytilid species, Mytilus galloprovincialis Lamarck, 1819, were encountered within the St Thomas' Bay community.
In a few places, like Spinola Bay, Manoel Island and the Msida marina (Table 1), high populations of Ostreola stentina (Payraudeau, 1826) seem to inhibit clustering of B. pharaonis, creating ‘breaks’ in what would otherwise be a complete stretch of Brachidontes-colonized coast. A wooden boat observed at the Spinola inlet yielded a vast population of O. stentina yet only a single, mature, individual of B. pharaonis. Two main reasons are postulated for this competitive exclusion—the size and complex nature of O. stentina clusters make it more difficult for larva of B. pharaonis to settle on the substratum, while their larger and more extensive filter-feeding apparatus may competitively inhibit the invasive species from obtaining the required nutrition. In addition, O. stentina is seemingly less selective as regards substratum (colonizing concrete walls more frequently) and is probably much more resistant to polluted water, such as in the innermost parts of the Msida marina, where it reaches very high population densities. In Sliema, areas containing juveniles of another lessepsian migrant bivalve Pinctada radiata (Leach, 1814) lacked any specimens of the mytilid.
The physical conditions in the sheltered area near Manoel Island, which is lacking in B. pharaonis, are similar to others where large densities of the species may be found. The bivalve's absence may be due to the extensive populations of Anemonia viridis Forsskål, 1775 which may preclude the larvae from settling.
A gastropod which may actively inhibit the settlement of B. pharaonis larvae is the predatory Pisania striata (Gmelin, 1791), whose presence in the current study mostly coincided with the absence or a relative scarcity of the mytilid. On the other hand, Osilinus turbinatus (Born, 1778), Muricopsis cristata (Brocchi, 1814), Cerithium lividulum Risso, 1826, patellids, chitons, and Chthamalus spp. were very frequently observed inhabiting the same areas as B. pharaonis. Bonnici (unpublished data) reports the following molluscan species as being exclusively reported from B. pharaonis beds (i.e. these species were missing from reference locations where B. pharaonis was only present as sparse individuals and not as beds): O. turbinatus, Ocinebrina edwardsii (Payraudeau, 1826) and Patella spp., together with Podocerus sp., nereids, nemerteans and sabellids.
A variety of macrofloral species characterized shores with B. pharaonis, with the main dominant species being Corallina elongata Ellis & Solander, Cladophora prolifera (Roth) Kützing, Ceramium ciliatum (Ellis) Ducluzeau, Ulva/Enteromorpha spp. and Jania rubens (Linné) Lamouroux along sheltered and polluted shores, and Cystoseira spp., Padina pavonica (Linné) Thivy, Sargassum vulgare C. Agardh and Dictyopteris polypodioides (De Candolle) Lamouroux along more exposed shores.
Besides ecological interactions, tolerance to the prevailing physical conditions obviously shape the current distribution of B. pharaonis, with exposure to wave action and phytoplankton quantity seemingly being the most important distribution determinants for the species. In fact, according to CIESM (Reference Zenatos, Gafas, Russo and Templado2003) ‘its abundance seems to be negatively associated with wave exposure'. The main repositories of B. pharaonis individuals recorded in the current study were sheltered areas, mainly within large embayments (Marsaxlokk Bay, Grand Harbour, Marsamxett Harbour and Salina), characterized by significant levels of water pollution in view of their popularity for shipping and recreational vessel berthing. Large breaks in the distribution of B. pharaonis have been recorded along the most exposed shores of the island of Malta (e.g. Pembroke—refer to Table 2). This finding is consistent with that of Safriel & Sasson-Frostig (Reference Safriel and Sasson-Frostig1988) but contradicts that reported by Rilov et al. (Reference Rilov, Benayahu and Gasith2004), who report extensive beds of B. pharaonis from wave-exposed stretches of the Israeli coastline. Bonnici (unpublished data) reports no significant variation between wave exposure values for locations with mussel beds formations and with sparse individuals.
ACKNOWLEDGEMENTS
The authors are grateful to the following individuals for assistance in work related to this study: Mr Edwin Lanfranco (identification of algal specimens), Mr Arnold Sciberras and Mr Jeffrey Sciberras (sampling duties), and Mr Charles Cachia and Mr Constantine Mifsud (provision of local records of the species).
